A computer system can be configured to run a variety of different applications. In today's fast cycle world, it is increasingly common to configure a computer system to also run different versions of the same application. Application files typically are stored on storage media directly attached to the computer system, however, they can also be located on network attached storage media when operating in a networked, or distributed computer environment.
In a Graphical User Interface (GUI) environment, the information required to launch an application is normally captured and stored as an icon or a shortcut. Traditionally, the user launches a particular application by selecting the name of the shortcut and instructing the operating system to execute it. The application selection process includes reading and parsing a directory or file names by the user and knowledge of the installed software. The present invention provides a solution that allows a self learning algorithm to identify and correctly launch the requisite application with minimal user interface by the use of a caching technique.
On the other hand, a network computing architecture includes a distributed computing environment, which generally comprises one or more server computers, one or more client PC's and may include one or more mainframe hosts. All computers are linked to one another through a network, which may be a local area network (LAN) running on an ethernet link, or a wide area network (WAN). The hard disk drive of each computer coupled to the network stores files in their structures. These files may or may not be shared with other computers in the network depending upon the permission granted by the network administrator. Where a large number of computers participate in such networks, it is essential to automate as much of the processes involved in distributed computing as possible. The present invention solves one of those critical processes and involves automating the search for files distributed throughout the network.